Seminars

 

in

 

Cell

 

&

 

Developmental

 

Biology

 

87

 

(2019)

 

125–144

Contents

 

lists

 

available

 

at

 

ScienceDirect

Seminars

 

in

 

Cell

 

&

 

Developmental

 

Biology

j

 

o u r n a

 

l

 

h

 

o

 

m e

 

p a g e :

 

w w w . e l s e v i e r . c o m / l o c a t e / s e m c d b

Review

Planarian

 

regeneration

 

as

 

a

 

model

 

of

 

anatomical

 

homeostasis:

 

Recent

progress

 

in

 

biophysical

 

and

 

computational

 

approaches

Michael

 

Levin a,b,∗,

 

Alexis

 

M.

 

Pietak a,

 

Johanna

 

Bischof a,b

a Allen

 

Discovery

 

Center

 

at

 

Tufts

 

University,

 

Medford,

 

MA

 

02155,

 

United

 

States

b Biology

 

Department,

 

Tufts

 

University,

 

Medford,

 

MA

 

02155,

 

United

 

States

a

 

r

 

t

 

i

 

c

 

l

 

e

 

i

 

n

 

f

 

o

Article

 

history:

Received

 

5

 

January

 

2018

Received

 

in

 

revised

 

form

 

3

 

April

 

2018

Accepted

 

6

 

April

 

2018

Available

 

online

 

1

 

May

 

2018

Keywords:
Planaria
Dugesia

 

japonica

Regeneration
Patterning
Morphostasis

a

 

b

 

s

 

t

 

r

 

a

 

c

 

t

Planarian

 

behavior,

 

physiology,

 

and

 

pattern

 

control

 

offer

 

profound

 

lessons

 

for

 

regenerative

 

medicine,

evolutionary

 

biology,

 

morphogenetic

 

engineering,

 

robotics,

 

and

 

unconventional

 

computation.

 

Despite

recent

 

advances

 

in

 

the

 

molecular

 

genetics

 

of

 

stem

 

cell

 

differentiation,

 

this

 

model

 

organism’s

 

remark-

able

 

anatomical

 

homeostasis

 

provokes

 

us

 

with

 

truly

 

fundamental

 

puzzles

 

about

 

the

 

origin

 

of

 

large-scale

shape

 

and

 

its

 

relationship

 

to

 

the

 

genome.

 

In

 

this

 

review

 

article,

 

we

 

first

 

highlight

 

several

 

deep

 

mysteries

about

 

planarian

 

regeneration

 

in

 

the

 

context

 

of

 

the

 

current

 

paradigm

 

in

 

this

 

field.

 

We

 

then

 

review

 

recent

progress

 

in

 

understanding

 

of

 

the

 

physiological

 

control

 

of

 

an

 

endogenous,

 

bioelectric

 

pattern

 

memory

that

 

guides

 

regeneration,

 

and

 

how

 

modulating

 

this

 

memory

 

can

 

permanently

 

alter

 

the

 

flatworm’s

 

target

morphology.

 

Finally,

 

we

 

focus

 

on

 

computational

 

approaches

 

that

 

complement

 

reductive

 

pathway

 

analy-

sis

 

with

 

synthetic,

 

systems-level

 

understanding

 

of

 

morphological

 

decision-making.

 

We

 

analyze

 

existing

models

 

of

 

planarian

 

pattern

 

control

 

and

 

highlight

 

recent

 

successes

 

and

 

remaining

 

knowledge

 

gaps

 

in

 

this

interdisciplinary

 

frontier

 

field.

©

 

2018

 

Elsevier

 

Ltd.

 

All

 

rights

 

reserved.

Contents

1.

 

Introduction

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

126

1.1.

 

A

 

primer

 

on

 

planarians’

 

functional

 

features

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

126

1.2.

 

Fundamental

 

knowledge

 

gaps .

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

126

1.3.

 

Perspective

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

128

2.

 

Physiological

 

controls

 

of

 

patterning .

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

. 128

2.1.

 

Prediction

 

1:

 

ion

 

channels

 

and

 

voltage

 

gradients

 

are

 

involved

 

in

 

planarian

 

patterning

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

128

2.2.

 

Prediction

 

2:

 

neurotransmitters

 

are

 

involved

 

in

 

planarian

 

patterning

 

control

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

130

2.3.

 

Prediction

 

3:

 

anatomical

 

outcome

 

and

 

genetic

 

default

 

can

 

diverge

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

130

2.4.

 

Prediction

 

4:

 

pattern

 

memory

 

can

 

be

 

over-written

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

133

2.5.

 

Summary:

 

physiological

 

controls

 

of

 

growth

 

and

 

form

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

134

3.

 

Computational

 

approaches

 

to

 

an

 

integrative

 

understanding

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

134

3.1.

 

Current

 

state

 

of

 

the

 

art

 

in

 

understanding

 

regenerative

 

dynamics:

 

gradients

 

and

 

beyond

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

135

3.2.

 

Advances

 

in

 

modeling

 

and

 

simulation:

 

testing

 

available

 

models

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

135

3.3.

 

Tools

 

for

 

model

 

discovery

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

137

4.

 

Conclusion

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

138

Acknowledgements

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

138

References

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

.

 

138

∗ Corresponding

 

author

 

at:

 

Allen

 

Discovery

 

Center

 

at

 

Tufts

 

University,

 

200

 

College

Avenue,

 

Medford,

 

MA

 

02155,

 

United

 

States.

E-mail

 

address:

 

michael.levin@tufts.edu

 

(M.

 

Levin).

https://doi.org/10.1016/j.semcdb.2018.04.003
1084-9521/©

 

2018

 

Elsevier

 

Ltd.

 

All

 

rights

 

reserved.

126

 

M.

 

Levin

 

et

 

al.

 

/

 

Seminars

 

in

 

Cell

 

&

 

Developmental

 

Biology

 

87

 

(2019)

 

125–144

1.

 

Introduction

1.1.

 

A

 

primer

 

on

 

planarians’

 

functional

 

features

Planarian

 

flatworms

 

are

 

free-living

 

bilaterian

 

organisms

 

with

 

a

complex

 

set

 

of

 

organ

 

systems

 

and

 

cell

 

types

 

[1,2].

 

Planaria

 

possess

rich

 

behavioral

 

repertoires

 

that

 

include

 

sensing

 

of

 

a

 

wide

 

range

 

of

environmental

 

cues

 

ranging

 

from

 

chemicals

 

[3]

 

to

 

gravity

 

[4],

 

to

weak

 

gamma

 

radiation

 

[5].

 

Though

 

planaria

 

have

 

a

 

true

 

brain

 

[6,7],

the

 

real

 

wonder

 

of

 

this

 

remarkable

 

model

 

system

 

is

 

revealed

 

most

clearly

 

when

 

one

 

considers

 

the

 

robust

 

decision-making

 

capabili-

ties

 

of

 

its

 

individual

 

somatic

 

pieces

 

[8].

 

Amputated

 

fragments

 

of

planaria

 

regenerate

 

into

 

a

 

complete

 

worm,

 

growing

 

precisely

 

what

is

 

missing

 

–

 

no

 

more

 

and

 

no

 

less

 

–

 

with

 

events

 

at

 

the

 

wound

 

edge

coordinated

 

tightly

 

with

 

body-wide

 

remodeling

 

of

 

the

 

intact

 

tis-

sue

 

to

 

ensure

 

that

 

a

 

perfectly

 

proportioned

 

animal

 

results

 

within

 

a

week

 

after

 

it

 

is

 

cut

 

along

 

any

 

plane

 

[9].

 

Every

 

piece

 

of

 

a

 

planarian

 

is

able

 

to

 

grow

 

and

 

remodel

 

towards

 

its

 

species-specific

 

large-scale

pattern,

 

stopping

 

precisely

 

when

 

that

 

specific

 

anatomical

 

pattern

(the

 

target

 

morphology)

 

is

 

achieved.

 

This

 

is

 

most

 

obvious

 

in

 

the

head-tail

 

polarity

 

of

 

their

 

primary

 

axis:

 

like

 

a

 

bar

 

magnet

 

cut

 

into

pieces

 

(Fig.

 

1

 

A),

 

each

 

piece

 

determines

 

head/tail

 

identity

 

at

 

the

wound

 

based

 

on

 

an

 

invariant

 

axial

 

polarity,

 

regenerating

 

into

 

a

worm

 

with

 

one

 

head

 

and

 

one

 

tail.

Planarians’

 

unparalleled

 

pattern

 

homeostasis

 

–

 

the

 

ability

 

to

adjust

 

cellular

 

activity

 

to

 

a

 

large-scale

 

anatomical

 

specification

[10,11]

 

–

 

is

 

an

 

on-going

 

process,

 

not

 

an

 

injury-specific

 

one.

 

Starving

animals

 

will

 

shrink,

 

and

 

well-fed

 

animals

 

will

 

grow,

 

repeating

 

this

process

 

ad

 

infinitum

 

while

 

maintaining

 

correct

 

body

 

proportions

[12].

 

Because

 

of

 

this

 

capability,

 

planaria

 

have

 

conquered

 

aging.

While

 

single

 

cells

 

senesce

 

and

 

die

 

within

 

relatively

 

short

 

time

spans

 

(

∼1

 

month)

 

[13–18],

 

the

 

animal

 

regenerates

 

tissues

 

con-

tinuously;

 

planaria

 

like

 

D.

 

japonica

 

appear

 

to

 

be

 

immortal

 

at

 

the

level

 

of

 

the

 

individual.

 

Planaria

 

are

 

the

 

champions

 

of

 

life-long

 

func-

tional

 

health

 

and

 

damage

 

repair.

 

However,

 

unlike

 

simpler

 

forms

of

 

life

 

such

 

as

 

bacteria

 

or

 

Hydra,

 

planaria

 

combine

 

these

 

abilities

with

 

significant

 

cognitive

 

potential

 

–

 

they

 

can

 

learn

 

in

 

instrumen-

tal

 

and

 

classical

 

conditioning

 

contexts

 

[19–24].

 

In

 

keeping

 

with

their

 

incredible

 

somatic

 

plasticity,

 

they

 

can

 

transfer

 

information

between

 

the

 

body

 

and

 

the

 

brain

 

[25–28],

 

and

 

reprogram

 

cancerous

tissue

 

when

 

regeneration

 

is

 

activated

 

[29,30].

Planaria

 

serve

 

as

 

a

 

proof-of-principle

 

–

 

showing

 

what

 

is

 

possible

for

 

biological

 

systems

 

to

 

achieve.

 

Therefore,

 

truly

 

understanding

planaria

 

would

 

crack

 

open

 

most

 

of

 

the

 

pressing

 

problems

 

in

 

the

fields

 

of

 

aging,

 

regenerative

 

medicine,

 

cancer,

 

and

 

primitive

 

cog-

nition.

 

Beyond

 

biology,

 

they

 

also

 

serve

 

as

 

a

 

design

 

challenge

 

for

fields

 

such

 

as

 

morphogenetic

 

engineering

 

[31,32],

 

unconventional

computation

 

[33–35],

 

and

 

soft

 

robotics

 

[36,37],

 

which

 

look

 

to

 

the

biological

 

world

 

for

 

clues

 

as

 

to

 

designing

 

robust,

 

parallel,

 

fault-

tolerant

 

systems.

 

The

 

planarian’s

 

ability

 

to

 

integrate

 

patterning

 

and

functional

 

control

 

is

 

a

 

superb

 

example

 

of

 

morphological

 

compu-

tation

 

[38]

 

and

 

distributed

 

decision-making

 

[39,40].

 

The

 

dynamic

adaptability

 

of

 

planaria

 

dwarfs

 

engineers’

 

efforts

 

on

 

this

 

front

 

–

 

no

human-designed

 

systems

 

even

 

come

 

close

 

to

 

their

 

capabilities

 

yet.

1.2.

 

Fundamental

 

knowledge

 

gaps

“No

 

paradox,

 

no

 

progress.”

 

-

 

Niels

 

Bohr

As

 

befits

 

a

 

model

 

system

 

with

 

fundamental

 

lessons

 

to

 

teach,

planaria

 

starkly

 

reveal

 

the

 

large

 

gaps

 

in

 

current

 

knowledge.

 

While

much

 

excellent

 

work

 

has

 

drilled

 

deeply

 

into

 

the

 

molecular

 

genetics

of

 

stem

 

cell

 

regulation

 

[41,42]

 

and

 

specific

 

pathways

 

necessary

 

for

regeneration

 

[43,44],

 

we

 

are

 

still

 

largely

 

missing

 

an

 

understand-

ing

 

of

 

the

 

dynamics

 

that

 

are

 

sufficient

 

to

 

build

 

an

 

organism

 

with

the

 

observed

 

remodeling

 

and

 

repair

 

functions.

 

Here,

 

we

 

argue

 

that

some

 

of

 

the

 

most

 

fascinating

 

aspects

 

of

 

this

 

field

 

concern

 

not

 

only

the

 

mechanisms

 

involved

 

at

 

the

 

cellular

 

and

 

molecular

 

levels,

 

but

also

 

the

 

algorithms

 

that

 

harness

 

individual

 

cell

 

behaviors

 

toward

specific

 

large-scale

 

anatomical

 

outcomes.

 

We

 

begin

 

by

 

highlight-

ing

 

a

 

few

 

of

 

the

 

mysteries

 

that

 

still

 

confront

 

us

 

despite

 

the

 

recent

advances

 

in

 

planarian

 

molecular

 

biology.

Establishing

 

correct

 

anterior-posterior

 

(AP)

 

polarity

 

involves

making

 

sure

 

every

 

fragment

 

cut

 

from

 

a

 

planarian

 

ends

 

up

 

with

exactly

 

one

 

head

 

and

 

one

 

tail,

 

at

 

the

 

appropriate

 

locations

 

(set-

ting

 

aside

 

for

 

now

 

the

 

issue

 

of

 

orienting

 

and

 

scaling

 

them

 

correctly,

giving

 

them

 

the

 

right

 

shape,

 

and

 

stopping

 

when

 

they

 

are

 

done).

Planaria

 

axially

 

pattern

 

each

 

fragment

 

no

 

matter

 

where

 

it

 

came

from

 

in

 

the

 

parent

 

structure

 

(Fig.

 

1A’)

 

–

 

the

 

challenging

 

nature

 

of

this

 

process

 

becomes

 

clear

 

when

 

one

 

considers

 

a

 

bisected

 

worm

(Fig.

 

1A”).

 

One

 

wound

 

site

 

will

 

make

 

a

 

head

 

while

 

the

 

other

 

will

make

 

a

 

tail;

 

the

 

cells

 

in

 

those

 

two

 

wound

 

sites

 

were

 

adjacent

 

neigh-

bors

 

(before

 

an

 

arbitrary

 

cut

 

separated

 

them)

 

and

 

yet

 

go

 

on

 

to

 

make

radically

 

different

 

anatomical

 

structures.

 

A

 

fundamental

 

challenge

is

 

to

 

understand

 

the

 

decision-making

 

process

 

that

 

occurs

 

at

 

each

wound

 

site

 

–

 

what

 

are

 

the

 

cells

 

measuring,

 

and

 

from

 

how

 

far

 

away,

to

 

determine

 

anatomical

 

identity

 

of

 

the

 

new

 

tissue

 

to

 

be

 

made?

This

 

is

 

difficult

 

for

 

simple

 

gradient

 

models

 

to

 

explain

 

(as

 

the

 

wound

cells

 

had

 

identical

 

positional

 

information

 

before

 

the

 

cut,

 

Fig.

 

1A”,

green

 

circles);

 

‘head-inhibits-head’

 

models

 

of

 

this

 

phenomenon

 

run

afoul

 

of

 

the

 

fact

 

that

 

a

 

brain

 

has

 

no

 

trouble

 

building

 

another

 

brain

right

 

next

 

to

 

it,

 

if

 

the

 

head

 

is

 

amputated

 

longitudinally,

 

as

 

well

as

 

being

 

unable

 

to

 

explain

 

axial

 

patterning

 

in

 

mid-fragments

 

with

open

 

wounds

 

on

 

both

 

sides.

The

 

second

 

puzzle

 

concerns

 

the

 

conspicuous

 

lack

 

of

 

stable

lines

 

of

 

planarian

 

patterning

 

mutants.

 

Most

 

other

 

model

 

organ-

isms

 

–

 

Drosophila,

 

C.

 

elegans,

 

chick,

 

mouse,

 

zebrafish,

 

and

 

humans

–

 

all

 

offer

 

stably

 

transmitted

 

lines

 

with

 

characteristic

 

and

 

strik-

ing

 

anatomical

 

deviations

 

from

 

wild-type

 

(patterning

 

mutants)

that

 

can

 

be

 

studied

 

to

 

forge

 

a

 

link

 

between

 

genetics

 

and

 

anatomy

(central

 

to

 

understanding

 

the

 

genotype-phenotype

 

relationship

 

in

evolution).

 

It

 

is

 

rarely

 

mentioned

 

in

 

our

 

field

 

that,

 

despite

 

almost

120

 

years

 

of

 

every

 

conceivable

 

experiment

 

including

 

irradiation,

with

 

the

 

exception

 

of

 

a

 

physiologically-induced

 

double-headed

 

line

described

 

below,

 

there

 

are

 

no

 

planarian

 

mutant

 

lines

 

with

 

patterns

(in

 

shape,

 

number,

 

or

 

placement

 

of

 

specific

 

organs)

 

that

 

differ

 

from

the

 

standard

 

species-specific

 

target

 

morphology.

 

Why?

The

 

answer

 

might

 

be

 

linked

 

to

 

the

 

third

 

puzzle:

 

their

 

remark-

able

 

anatomical

 

stability

 

in

 

the

 

face

 

of

 

a

 

highly

 

variable

 

genome.

Weissmann’s

 

Barrier

 

(Fig.

 

1B,

 

B’)

 

does

 

not

 

apply

 

to

 

species

 

like

 

D.

japonica

 

which

 

reproduce

 

primarily

 

asexually

 

by

 

fission

 

and

 

regen-

eration.

 

In

 

this

 

case,

 

any

 

mutation

 

that

 

does

 

not

 

kill

 

a

 

neoblast

 

is

propagated

 

to

 

the

 

next

 

generation

 

and

 

can

 

expand

 

into

 

a

 

clonal

line

 

(Fig.

 

1B”,

 

B”’).

 

This

 

will

 

be

 

especially

 

true

 

of

 

the

 

inevitable

dominant

 

mutations,

 

which

 

increase

 

relative

 

cellular

 

fitness

 

and

could

 

become

 

something

 

akin

 

to

 

cancer

 

stem

 

cells.

 

This

 

somatic

inheritance

 

(Fig.

 

1C–C”’)

 

is

 

predicted

 

to

 

generate

 

extremely

 

diver-

gent

 

genomes

 

over

 

time

 

[45];

 

indeed

 

not

 

only

 

are

 

some

 

planarian

species

 

mixoploid

 

(not

 

every

 

cell

 

has

 

the

 

same

 

chromosome

 

num-

ber)

 

[46],

 

but

 

they

 

also

 

accumulate

 

immense

 

amount

 

of

 

change:

 

up

to

 

74%

 

in

 

protein-coding

 

genes

 

[47].

 

Indeed,

 

large

 

numbers

 

of

 

muta-

tions

 

have

 

been

 

found

 

both

 

outside

 

and

 

within

 

gene-coding

 

regions,

including

 

many

 

amino

 

acid

 

substitutions

 

and

 

non-synonymous

SNPs.

 

Furthermore,

 

a

 

recent

 

analysis

 

of

 

the

 

genome

 

of

 

the

 

pla-

naria

 

Schmidtea

 

mediterranea

 

revealed

 

that

 

many

 

essential

 

genes

appear

 

to

 

be

 

missing

 

from

 

the

 

genome,

 

including

 

components

 

of

many

 

core

 

pathways

 

ranging

 

from

 

cell

 

division,

 

to

 

DNA

 

repair

and

 

metabolism

 

[48].

 

And

 

yet,

 

these

 

animals

 

regenerate

 

under

control

 

conditions

 

with

 

100%

 

anatomical

 

fidelity,

 

making

 

perfect

planaria

 

each

 

time

 

despite

 

their

 

messy

 

genomes.

 

How

 

is

 

it

 

possible

for

 

the

 

genome

 

to

 

accumulate

 

somatic

 

mutations

 

over

 

hundreds

of

 

millions

 

of

 

years

 

and

 

yet

 

maintain

 

perfect

 

anatomical

 

fidelity?

M.

 

Levin

 

et

 

al.

 

/

 

Seminars

 

in

 

Cell

 

&

 

Developmental

 

Biology

 

87

 

(2019)

 

125–144

 

127

Fig.

 

1.

 

Planarian

 

regeneration:

 

fundamental

 

puzzles

 

of

 

pattern

 

control.

(A)

 

Bar

 

magnets

 

illustrate

 

a

 

basic

 

property

 

of

 

polarity

 

re-scaling:

 

having

 

a

 

North

 

and

 

South

 

pole,

 

a

 

magnet

 

can

 

be

 

cut

 

into

 

pieces

 

and

 

each

 

piece

 

reorganizes

 

its

 

polarity

 

to

likewise

 

have

 

a

 

North

 

and

 

South

 

pole

 

by

 

orientating

 

small

 

magnetic

 

domains

 

into

 

large-scale

 

axial

 

patterning.

(A’)

 

Planarian’s

 

anterior-posterior

 

(AP)

 

axis

 

likewise

 

re-scales:

 

every

 

piece

 

cut

 

from

 

a

 

planarian

 

(D.

 

japonica

 

shown

 

here)

 

regenerates

 

a

 

head

 

and

 

tail

 

at

 

the

 

correct

 

end.

(A”)

 

Common

 

models

 

of

 

axial

 

patterning

 

postulate

 

a

 

chemical

 

gradient

 

that

 

indicates

 

positional

 

information

 

for

 

cells

 

along

 

the

 

AP

 

axis.

 

The

 

ability

 

of

 

cells

 

on

 

either

 

side

 

of

 

a

bisection

 

to

 

develop

 

distinct

 

anatomical

 

fates

 

(head

 

vs.

 

tail),

 

even

 

though

 

they

 

began

 

as

 

adjacent

 

neighbors

 

with

 

identical

 

positional

 

information

 

(green

 

circles),

 

suggests

the

 

need

 

for

 

long-range

 

communication

 

across

 

the

 

fragment

 

so

 

that

 

decisions

 

at

 

the

 

wound

 

could

 

be

 

made

 

based

 

on

 

the

 

rest

 

of

 

the

 

fragment.

128

 

M.

 

Levin

 

et

 

al.

 

/

 

Seminars

 

in

 

Cell

 

&

 

Developmental

 

Biology

 

87

 

(2019)

 

125–144

This

 

paradox

 

is

 

pointing

 

to

 

a

 

fundamental

 

lack

 

of

 

knowledge

 

of

“where

 

pattern

 

comes

 

from”

 

and

 

what

 

part

 

genetics

 

plays

 

in

 

the

process.

 

We

 

are

 

currently

 

investigating

 

this

 

issue

 

by

 

raising

 

gener-

ations

 

of

 

planaria

 

in

 

high

 

doses

 

of

 

various

 

mutagens,

 

to

 

fully

 

probe

the

 

stability

 

of

 

the

 

animals’

 

pattern

 

homeostasis

 

under

 

genetic

change.

The

 

fourth

 

puzzle

 

to

 

consider

 

is

 

a

 

thought

 

experiment.

 

Suppose

 

a

hybrid

 

was

 

made

 

between

 

planaria

 

with

 

two

 

distinct

 

head

 

shapes:

removing

 

half

 

of

 

the

 

neoblasts

 

of

 

one

 

species

 

of

 

planaria,

 

and

 

pop-

ulating

 

it

 

with

 

50%

 

of

 

the

 

neoblasts

 

from

 

another

 

species

 

(Fig.

 

1D).

After

 

the

 

transplanted

 

neoblasts

 

adjust

 

to

 

their

 

new

 

home,

 

the

 

head

is

 

then

 

amputated

 

(Fig.

 

1D’,

 

D”).

 

The

 

neoblasts

 

will

 

start

 

building

a

 

new

 

head

 

–

 

but

 

what

 

shape

 

will

 

form

 

(Fig.

 

1D”’)?

 

Is

 

one

 

set

 

of

neoblasts

 

dominant

 

over

 

the

 

other,

 

will

 

the

 

resulting

 

head

 

be

 

an

intermediate

 

shape,

 

or

 

will

 

the

 

remodeling

 

never

 

cease,

 

as

 

each

set

 

of

 

neoblasts

 

is

 

never

 

given

 

the

 

“stop”

 

signal

 

derived

 

from

 

hav-

ing

 

a

 

complete,

 

normally-shaped

 

head

 

appropriate

 

to

 

its

 

species?

The

 

value

 

of

 

this

 

thought

 

experiment

 

is

 

not

 

so

 

much

 

only

 

in

 

the

answer,

 

but

 

in

 

the

 

fact

 

that

 

we

 

currently

 

have

 

no

 

models

 

with

 

which

we

 

could

 

make

 

a

 

prediction

 

about

 

this

 

basic

 

scenario.

 

The

 

com-

bined

 

communities’

 

best

 

quantitative

 

models,

 

including

 

extensive

RNA-seq

 

profiling,

 

do

 

not

 

constrain

 

the

 

outcome

 

of

 

such

 

experi-

ments

 

–

 

they

 

are

 

silent

 

about

 

shape

 

and

 

what

 

determines

 

it,

 

and

it

 

is

 

not

 

yet

 

clear

 

how

 

much

 

of

 

the

 

necessary

 

information

 

is

 

gen-

erated

 

by

 

the

 

soma

 

and

 

how

 

much

 

by

 

the

 

neoblasts.

 

Thus,

 

our

field

 

lacks

 

theoretical

 

frameworks

 

for

 

thinking

 

about

 

large-scale

states

 

(e.g.,

 

head

 

shape)

 

as

 

targets

 

and

 

stop

 

conditions

 

for

 

pattern

homeostasis

 

loops

 

that

 

regulate

 

cell

 

behavior.

 

This

 

is

 

a

 

necessity

 

if

we

 

are

 

to

 

understand

 

how

 

cells

 

know

 

what

 

to

 

build

 

and

 

when

 

to

stop.

1.3.

 

Perspective

We

 

highlight

 

these

 

puzzles

 

as

 

examples

 

of

 

questions

 

that

emphasize

 

major

 

knowledge

 

gaps,

 

and

 

thus

 

opportunities,

 

in

 

this

field.

 

We

 

argue

 

that

 

significant

 

new

 

conceptual

 

and

 

technological

approaches

 

must

 

be

 

added

 

to

 

the

 

highly

 

successful

 

mainstream

efforts

 

to

 

understand

 

specific

 

gene

 

products

 

and

 

cell-level

 

phe-

notypes.

 

Here,

 

we

 

use

 

planarian

 

regeneration

 

as

 

a

 

lens

 

through

which

 

we

 

can

 

view

 

several

 

broad

 

issues

 

of

 

biological

 

computation,

pattern

 

control,

 

and

 

the

 

genotype-phenotype

 

relation.

 

Advances

in

 

biochemical

 

and

 

genetic

 

controls

 

of

 

regeneration

 

have

 

been

expertly

 

reviewed

 

elsewhere

 

[44,49–51].

 

Focusing

 

on

 

synthesis

 

and

understanding

 

global

 

morphological

 

decision-making

 

in

 

this

 

model

system,

 

we

 

review

 

recent

 

developments

 

in

 

this

 

field

 

that

 

have

 

the

potential

 

to

 

drive

 

progress

 

on

 

fundamental

 

aspects

 

of

 

the

 

origin

and

 

control

 

of

 

growth

 

and

 

form.

 

These

 

include

 

(1)

 

endogenous

physiological

 

signals

 

that

 

underlie

 

pattern

 

control:

 

bioelectric

 

and

neurotransmitter-mediated

 

signaling

 

mechanisms

 

in

 

numerous

cell

 

types,

 

which

 

enable

 

long-range

 

coordination

 

and

 

morpholog-

ical

 

decision-making

 

by

 

cell

 

collectives

 

[40];

 

and

 

(2)

 

advances

 

in

computational

 

modeling

 

and

 

automated

 

model

 

discovery,

 

which

are

 

helping

 

to

 

understand

 

the

 

algorithms

 

by

 

which

 

planarian

 

shape

is

 

controlled.

2.

 

Physiological

 

controls

 

of

 

patterning

Regeneration

 

of

 

significant

 

injury

 

requires

 

rebuilding

 

structures

that

 

are

 

properly

 

coordinated

 

in

 

position,

 

orientation,

 

and

 

size

with

 

the

 

large-scale

 

anatomy

 

of

 

the

 

remaining

 

body,

 

which

 

implies

that

 

cells

 

need

 

non-local

 

information

 

to

 

make

 

patterning

 

deci-

sions

 

(Fig.

 

1A’).

 

The

 

ability

 

to

 

re-create

 

the

 

same

 

structure

 

time

and

 

again

 

can

 

be

 

understood

 

as

 

a

 

“pattern

 

memory”

 

[39,52],

 

while

the

 

ability

 

to

 

reach

 

the

 

same

 

correct

 

pattern

 

from

 

different

 

initial

starting

 

conditions

 

(location

 

and

 

extent

 

of

 

damage)

 

implies

 

a

 

robust

goal-directed

 

process.

 

One

 

way

 

to

 

think

 

about

 

the

 

remarkable

decision-making

 

properties

 

of

 

planarian

 

tissues

 

is

 

to

 

consider

 

the

algorithms

 

and

 

molecular

 

mechanisms

 

exploited

 

by

 

brains

 

–

 

our

best

 

example

 

of

 

biological

 

systems

 

that

 

implement

 

memory,

 

dis-

tributed

 

processing,

 

decision-making,

 

and

 

flexible

 

goal-achieving

cell

 

networks.

 

Interestingly,

 

planaria

 

were

 

one

 

of

 

the

 

earliest

model

 

systems

 

in

 

which

 

data

 

suggested

 

roles

 

for

 

ion-

 

and

 

voltage-

mediated

 

processes

 

in

 

guiding

 

regeneration

 

[53–57].

 

Since

 

then,

advances

 

in

 

developmental

 

bioelectricity

 

(reviewed

 

in

 

[58,59])

have

 

driven

 

hypotheses

 

about

 

the

 

role

 

of

 

ion

 

channels

 

in

 

guiding

cell

 

behavior,

 

and

 

more

 

broadly,

 

about

 

the

 

relationships

 

between

the

 

activity

 

of

 

multi-cellular

 

electric

 

circuits

 

and

 

developmental

morphospace

 

[60,61].

 

This

 

perspective

 

made

 

several

 

specific

 

and

counter-intuitive

 

predictions,

 

driving

 

new

 

experiments

 

that

 

com-

plement

 

the

 

biochemical/genetic

 

research

 

programs.

 

Recent

 

work

has

 

tested

 

some

 

of

 

these

 

predictions,

 

uncovering

 

novel

 

biology

 

in

the

 

planarian

 

model

 

system.

2.1.

 

Prediction

 

1:

 

ion

 

channels

 

and

 

voltage

 

gradients

 

are

 

involved

in

 

planarian

 

patterning

Neurons

 

compute

 

and

 

transmit

 

information

 

long-distance

 

by

virtue

 

of

 

electrical

 

signaling.

 

It

 

is

 

now

 

well-recognized

 

that

 

even

neural

 

computations

 

can

 

involve

 

graded

 

(not

 

spiking)

 

potentials

[62–66],

 

which

 

are

 

similar

 

to

 

the

 

slower

 

non-neural

 

bioelectric

events

 

that

 

operate

 

in

 

regeneration

 

and

 

development

 

(Fig.

 

2A).

Could

 

non-neural

 

somatic

 

tissues

 

be

 

performing

 

information-

processing

 

tasks

 

by

 

exploiting

 

ion

 

flows,

 

albeit

 

in

 

different

 

ways

and

 

on

 

different

 

timescales

 

than

 

brains?

 

While

 

developmental

 

biol-

ogists

 

are

 

not

 

yet

 

used

 

to

 

thinking

 

of

 

non-neural

 

tissues

 

making

decisions,

 

the

 

emerging

 

field

 

of

 

primitive

 

cognition

 

[67,68]

 

and

 

the

recent

 

data

 

on

 

the

 

phylogeny

 

of

 

ion

 

channel

 

and

 

neurotransmit-

ter

 

signals

 

[69–71]

 

have

 

highlighted

 

the

 

fact

 

that

 

brains

 

did

 

not

invent

 

their

 

tricks

 

de

 

novo

 

–

 

the

 

basic

 

machinery

 

of

 

bioelectrical

computation

 

(Fig.

 

2A’)

 

was

 

present

 

very

 

early

 

on

 

in

 

evolution

 

and

is

 

ubiquitous

 

in

 

animal,

 

plant,

 

and

 

fungal

 

bodies.

 

Even

 

before

 

mul-

ticellularity,

 

cells

 

were

 

using

 

ion

 

channels

 

and

 

neurotransmitters

to

 

process

 

information

 

and

 

communicate;

 

from

 

bacteria

 

and

 

fungi

to

 

the

 

earliest

 

metazoans,

 

cells

 

exploited

 

ion

 

currents

 

to

 

regulate

individual

 

and

 

group

 

behaviors

 

(data

 

in

 

vertebrate

 

as

 

well

 

as

 

inver-

tebrate

 

systems

 

are

 

reviewed

 

in

 

[59,72–74]).

Research

 

into

 

bioelectric

 

aspects

 

of

 

patterning

 

has

 

a

 

rich

 

his-

tory

 

[75,76].

 

The

 

ideal

 

fit

 

of

 

bioelectric

 

circuits

 

to

 

the

 

control

 

of

regeneration

 

did

 

not

 

escape

 

prescient

 

early

 

workers

 

such

 

as

 

T.

 

H.

Morgan,

 

who

 

postulated

 

electrical

 

polarity

 

to

 

underlie

 

regenerative

axial

 

polarity

 

[77,78],

 

and

 

C.

 

M.

 

Child,

 

who

 

focused

 

on

 

physiologi-

(B,

 

C)

 

For

 

most

 

metazoans,

 

sexual

 

reproduction

 

results

 

in

 

Weissmann’s

 

barrier

 

–

 

somatic

 

mutations

 

(B)

 

do

 

not

 

persist

 

into

 

the

 

next

 

generation

 

(B’).

 

However,

 

planaria

 

such

 

as

D.

 

japonica

 

largely

 

reproduce

 

through

 

fission;

 

thus,

 

mutations

 

in

 

neoblasts

 

anywhere

 

in

 

the

 

worm

 

body

 

(B”)

 

that

 

do

 

not

 

kill

 

the

 

neoblast

 

persist

 

into

 

the

 

next

 

generation

 

upon

fission

 

(B”’).

 

Within

 

that

 

individual

 

(C),

 

the

 

progeny

 

of

 

the

 

neoblast

 

inherit

 

the

 

mutation(s),

 

which

 

spread

 

throughout

 

the

 

body

 

(C’)

 

and

 

are

 

propagated

 

to

 

both

 

offspring

 

of

 

the

next

 

fission

 

event

 

(C”).

 

This

 

gives

 

rise

 

to

 

a

 

fundamental

 

puzzle

 

about

 

the

 

relationship

 

between

 

genetics

 

and

 

patterning:

 

over

 

hundreds

 

of

 

millions

 

of

 

years,

 

planarian

 

lineages

have

 

accumulated

 

diverse

 

mutations

 

in

 

their

 

bodies;

 

despite

 

the

 

resulting

 

very

 

messy

 

genomes,

 

planaria

 

regenerate

 

with

 

100%

 

anatomical

 

fidelity

 

and

 

offer

 

no

 

genetic

 

lines

of

 

patterning

 

mutants.

(D)

 

A

 

thought

 

experiment

 

involving

 

two

 

different

 

species

 

of

 

flatworms

 

with

 

distinct

 

head

 

shapes,

 

illustrates

 

knowledge

 

gaps

 

in

 

the

 

field.

 

If

 

one

 

of

 

the

 

worms

 

is

 

irradiated

so

 

that

 

half

 

of

 

its

 

neoblasts

 

are

 

killed

 

(D’),

 

and

 

neoblasts

 

from

 

another

 

worm

 

type

 

are

 

injected

 

into

 

this

 

host

 

(D”),

 

what

 

kind

 

of

 

head

 

shape

 

will

 

this

 

hybrid

 

worm

 

regenerate

after

 

amputation

 

(D”’)?

 

Existing

 

models

 

do

 

not

 

make

 

a

 

prediction

 

as

 

to

 

whether

 

a

 

dominant

 

shape,

 

a

 

combination

 

shape,

 

or

 

continuous

 

remodeling

 

will

 

result.

Panels

 

B–C”’

 

were

 

made

 

by

 

Jeremy

 

Guay

 

of

 

Peregrine

 

Creative.

 

Panels

 

D-D”’

 

are

 

used

 

with

 

permission

 

from

 

[8].